Virtual Time Virtual Time and Global States of Distributed Systems Friedmann Mattern, 1989 The Model: An asynchronous distributed system = a set of processes having no shared memory, communicating by message transfer. Message delay > 0, but is not known in advance. A global observer – sees the global state at certain points in time. It can be said to take a snapshot of the global state. A local observer – (one of the processes in the system) sees the local state. Because of the asynchrony, a local observer can only gather local views to an approximate global view. This is a hard hazard for many management and control problems: Mutual exclusion, deadlock detection, distributed contracts, leader election, load sharing, checkpointing etc.

Solution Approaches 1.Simulating a synchronous system by an asynchronous one. This requires high overhead on global synchronization of each and every step. 2.Simulation of a global state. A snapshot, taken asynchronously, which is not necessarily correct for any specific point in time, but is in a way consistent with the local states of all processes. 3.A logical clock which is not global, but can be used to derive useful global information. The system works asynchronously, but the processes make sure to maintain their part of the clock.

Events An event is a change in the process state. An event happens instantly, it does not take time. A process is a sequence of events There are 3 types of events: 1.send event – causes a message to be sent 2.receive event – causes a message to be received 3.local event – only causes an internal change of state Events correspond to each other as follows: All events in the same process happen sequentially, one after the other. Each send event has a corresponding receive This allows us to define the happened before relation among events.

The Happened Before Relation We say that event e happened before event e (and denote it by e e or e < e) if one of the following properties holds: P1 P2 P3 e21 e11 e31 e22 e13 global time e12 e23 e32 e14 Program order:e13 < e14 Send-Receive: e23 < e12 Transitivity:e21 < e32 Processor Order: e precedes e in the same process Send-Receive: e is a send and e is the corresponding receive Transitivity: exists e s.t. e < e and e< e Example:

Independent/ Concurrent Events Two such diagrams are called equivalent when the happened before relation is the same in both. (When global time differs for certain events, think of processor execution as if it was a rubber band). Two events e, e are said to be independent or concurrent (denoted by e || e) if not e < e and not e < e. P1 P2 P3 e21 e11 e31 e22e23e24e25 e32e33e34 e12e13 global time

A cut C of a set of events E is the union set of the cut events set { ci } and all the events e E which precede any of the ci in program order. A cut is said to be consistent if also: e C and e<e always imply e C. If a receipt of a message is in a consistent cut – so is its sending. The opposite does not necessarily hold. Informally, a cut is consistent if when stretching the execution lines so that the cut events form a vertical line, there is no send-receive arrow which crosses the cut right to left. Cuts future past c1 P1 P2 P3 e11 e41 e42 e12 P4 e22e21 e31 c2 c3 c4 Ci are events of the cut.

Global States = Consistent Cuts If a cut is consistent then there is a real execution of the diagram in which all cut events happen at the same moment. Clearly, the view of a global observer at any point in time is necessarily consistent (a message that arrived was necessarily previously sent). Thus, a global view determines a consistent cut at any point in time. Any consistent cut may actually happen at a certain point during a real execution. It contains the local states of the processes when the cut event happens, and the set of messages that were already sent but not received. The Global Snapshot problem: Find an efficient protocol to compute consistent cuts. i.e. collect the local process states and the messages cutting the cut.

Virtual Time (Lamport, 1978) A logical clock is a function C:E T E – a set of events, C(e) timestamp of e T – a partially ordered set s.t. e<e C(e)<C(e) (the opposite not necessarily true, e.g. concurrent events.) Commonly, T=N, and there is a local clock Ci for each process Pi. The clocks perform the following protocol: 1.When a local event is executed in Pi: Ci := Ci + d (d>0) 2.Each message m, sent by event e = send(m), is time-stamped t(m) = C(e), and this timestamp is sent with the message. 3.When Pi receives a message with timestamp t: Ci: = max(Ci,t(m)) + d (d >0) Usually, d=1. However, d may change arbitrarily and dynamically (as long as it is positive), e.g., in an attempt to reflect actual time. The timestamp of e, C(e), is given after advancing the clock by d, accounted for e.

Logical Clocks Cntd. P1 P2 P3 e11e12 e13 e21e22 e31 C1=1 C1=2 C1=3 C2=1C2=2 C3=3 A problem: When the history of execution is read, only the timestamps of the events are known, and there is a loss of information: Suppose C(e) < C(e). We do know that e cannot have happened before e, so not (e<e). But we do not know whether e < e or e || e. Information whether the events are independent is lost (a global observer would know). Reason: When e and e are concurrent, then any of C(e) < C(e), C(e) < C(e), C(e) = C(e) may hold. Concurrency information is lost. Example:

A Vector Clock for a Global Observer P1 P2 P3 P cuts Vector of times as seen by a global observer

Approximating Global Vector Time Idea: lets compute for every process an approximation of the global observers vector clock. Have to show that approximation is good enough. Instead of integer clocks, each process Pi will maintain a vector clock Ci[]. Thus, timestamp C(e) of an event e at process Pi is Ci[] when e happens. The new protocol for maintaining the clocks is as follows: 1.When executing a local event in Pi: Ci[i] := Ci[i] Each message m, sent by event e = send(m), is time-stamped t(m) = C(e). 3.Upon receipt of a message m: for every j: Ci[j] := max{Ci[j],t(m)[j]} The time stamp of a received message contains information about other processes states. On receipt of a message, a process combines this information with what it already knows about the system state, and communicates it elsewhere, piggybacking on messages sent out.

Back to Example P1 P2 P3 P Vector of times as seen by a global observer C1 C2 C3 C t(m)= m m m Note: P3 receives ticking information from P2 even though they never communicate directly.

Actual Order of Independent Events not Important P1 P2 P3 P m m m What is the actual global order? Really dont Care! However global time = is impossible

An Invariant (*) Only Pi can advance his clock: Ci[i] ticks only at Pi. Thus, Pi always knows the up-to-date time of his local time. Thus, at any point in time, for any i,j, it holds that: Ci[i] >= Cj[i]

Vector Times can be Compared u and v concurrent: u happened before v: u did not happen after v: With vector times, timestamps can keep all the necessary information concerning the order of events. Claim: time stamp relations and happen before relations of corresponding events are equivalent.

Vector Times Relations Represent Order of Events Claim: for every e,e 1. C(e) < C(e) iff e < e 2. C(e) || C(e) iff e || e 3.If in addition e is known to have happened in process Pi then for every e then: C(e)[i] <= C(e)[i] iff e < e Proof: (For (1). Proofs of (2,3) are similar.) If e knows the local time of e (i.e., the vector component corresponding to the process where e happened is bigger in e than in e), then there must exist a sequence of time stamps from e to e which represents the happened before relation. If e<e then there exist a sequence of time stamps from e to e, and these time stamps monotonically grow in each component. The growth is strict in at least one component: that corresponding to es process.

Cuts Revisited Let X be a cut, {xi} cut events, a system of n processes. We say that the global time of X is tx = max(C(x1), …, C(xn)) (max is taken per entry). Example: Different cuts have same global Time However …

Different Consistent Cuts Have Different Global Times Claim: X is consistent iff tx = (C(x1)[1], …,C(xn)[n]) Proof: If X is consistent then its cut events xi can be viewed as if they occurred at the same moment. By the invariant (*) at that moment the claim holds. If X is not consistent, then there exists a message that was sent (say) from Pi after xi, and was received by Pj before xj. Let t be the timestamp on the message, then xi[i] < t[i] <= xj[i]. Thus tx > (C(x1)[1], …,C(xn)[n]).